Non-frost domestic refrigerator evaporators
The work carried out in CTTC on this applied topic has been the core of our fin-and-tube heat exchangers applied research line.
As a consequence of several years of academic research and industrial collaboration, an interesting combination of a distributed numerical model, experimental studies and CFD analysis has been developed.
Â
Distributed numerical model
The Group has been continuously developing a distributed numerical model to analyse compact heat exchangers (called CHESS), which was mainly centered on automobile radiators, water coolers, and evaporators/condensers for air-conditioning applications. The code has recently been adapted to household evaporators, with special emphasis on particular fin spacing distributions and frost formation.
The numerical method is based on the discretization of the heat exchanger in a set of macro control volumes around the tubes, where governing equations are applied to obtain the 3D temperature, velocity and pressure maps. For the airside, the humidity map / frost formation is also calculated.
The corresponding fins are modeled either by using analytical fin efficiency, or by an additional multidimensional numerical module (considering thermal bridges, transient term,…). The tubes are numerically analyzed considering the axial and peripheral heat transfer, and the contact with the fin array.
For the refrigerant side, a quasi-homogeneous two-phase flow model has been implemented to consider the evaporation process. The appropriate empirical information is imposed depending on the flow state and pattern.
Regarding the set of available results from the model, as a first output the model provides a set of overall values calculated at each time step, allowing for the analysis of the transient evolution of the evaporator (heat transfer, SHR, frost accumulation, etc.).
On the other hand, for each time step or for steady state calculations, a very detailed picture of the evaporator behavior is available from model results (3D solid temperature maps, profiles along refrigerant path, profiles along air flow path, etc.).
Â
Experimental set-up
An air loop has been constructed to test fin-and-tube evaporators within the very specific operating conditions found in this application (very low airflows, low temperature and humidity). The airflow inlet velocity profile has been conditioned by adequate screens and flow straighteners (and verified by hot-wire anemometry measurements). The airflow is circulated by means of a centrifugal blower and measured by a vortex flowmeter, taking values from about 15 to 60 m3/h. Temperature is controlled by means of an auxiliary air-to-liquid heat exchanger (connected to a secondary thermostatic bath), combined with an electric heater. The humidity is controlled by a high precision vapour injection system (reported injection precision from about 6 g/h). The duct is kept vertical in order to replicate exactly the position within the refrigerator (especially relevant with isobutane two-phase flow, where stratified flow pattern could have an impact). The air temperature is measured by two thermocouple (TC) grids (4×2 matrix) near the evaporator inlet/outlet duct sections, and by two thermoresistances (RTD) after inlet/outlet mixing sections. A radiation shielded TC is placed in each grid in order to confirm radiation negligible influence. The humidity is measured by two high quality capacitive sensors.
As stated before, the evaporator is placed inside the vertical duct, in a section where transparent insulation allows a qualitative analysis of the frost formation phenomena. Particularly in two-phase flow conditions, where one part of the refrigerant flow becomes superheated vapour, the tube temperature pattern strongly influences the frost formation distribution. In common tests, this transparent section are substituted by standard insulation to have better heat transfer conditioning.
CFD analysis through fin-and-tube core
The distributed model has been usually fed by open-literature heat transfer and friction correlations in order to predict accurately and without any experimentally based correction the numerical results. However, this approach is limited to available geometries and ranges and considering the excellent CFD background of the research Group, in-house CFD studies on the fin-and-tube geometries involved in household refrigerator evaporators are being carried out. After proper validation against published and own experimental results, the combination of CFD-based correlations with a very detailed distributed model results in a very powerful design tool.
As illustrative results, air temperature, velocity and pressure maps are depicted in the following figures for a geometry and flow range similar to those find in household refrigerator evaporators. The studies are nowadays focused in the development of advanced interpolation schemes to speed up conjugate heat transfer cases for non-uniform fin distributions.
